I quite like the Swiss model of democracy. Taleb argues that the focus on small matters and diffused democracy avoids centralization of power and provides a lot of their famous stability.
On the other hand, several Communists (in fact, the ones Zizek is arguing against) very much argue for exactly what you propose: focus on small matters and diffused democracy to avoid centralisation of power. The fact of "real Communism" having been or not having been "tried" is irrelevant to the questions concerning how such a society, if it is possible, ought to be organised. Thought-terminating cliches don't make for critical analysis.
[0] https://plato.stanford.edu/entries/marcuse/#FanUtoRatGra
You need to see how things actually work out in practice. And in practice Swiss democracy works very well whereas most central bureaucracies are soul killers.
In case you do not know, Swiss have a central bureaucracy. Not everything, heck most things are not decided by direct vote. Important things are though and the citizens can propose a referendum after the fact as well given enough buy in.
The problem with this kind of direct democracy is the same as with all diplomacy - astroturfing, focus groups, demagoguery and unintended consequences.
Focus groups make diffuse votes matter less despite being more prevalent. (Overwhelming majority in some cases.) Astroturfing is a kind of diffuse bribery. Unintended consequences is most often when people are presented with a package deal and do not dig deeply enough to figure out results. Finally demagoguery is usually by presenting a palatable but highly inferior option or by going for short term bandaid solutions.
The latter two are less relevant to a programming language project.
